Recent research on the prevalence of HIV infection among illicit drug users has shown that crack smokers have an unexpectedly high HIV seroprevalence. Interventions aimed at reducing sexual risk behavior among young adult crack smokers and other non-injection drug users (NIDUs) are urgently needed. We propose a randomized controlled trial of an innovative HIV risk-reduction co-educational intervention (COPE-Network), which incorporates cognitive-behavioral skills building and social influencing approaches. The primary aim is to reduce sexual risks associated with HIV seroconversion. To achieve this aim, we will recruit young sexually active 11W negative NIDUs (18-30 year old) in Harlem, New York City. We plan to enroll 300 NIDUs aged 18 to 30 into a randomized trial with two arms (COPE-Network versus a control condition; 150 index persons in each condition), as well as members of their sex/drug personal networks. The aims of the study are: (1) To assess the efficacy of the intervention on the primary sexual risk behavior outcomes (VEE, a weighted sexual risk index, and other measures such as proportion of condom-protected acts, and number of sex partners) among index subjects; (2) To assess the efficacy of the intervention on the primary sexual risk behavior outcomes among network subjects; (3) To explore associations between high risk sexual behaviors and risk configuration (the context in which high risk behaviors occur) among index and control subjects. Risk configurations will be assessed by considering the proportion of personal network members engaging in high-risk sexual behavior and illicit drug use; and (4) To gather descriptive epidemiological data on the prevalence and incidence of gonorrheal and Chlamydia infections among the index NIDUs enrolled in the clinical trial. Additionally, these data will be used to provide corollary evidence of an intervention effect by analyzing correlations with self-reported measures of sexual behavior.